Thomas Nolle: Lessons from Lucent’s 880 switch

Lucent Technologies Inc.’s announcement of its TMX 880 switch might be one of the pivotal events of what I hope will become known as the “telecom recovery period.” The product addresses many of the key issues the industry is facing now. One way or the other, the 880 is likely to touch every player in the market.

Voice profits are falling off rapidly for all carriers, and legacy data services such as ATM and frame relay offer relatively little incremental growth opportunity. IP is the future, but we’ve had so many false starts getting to profitable IP that Wall Street has about given up. That’s the first reason why the 880 is so important. By combining an IP-centric feature – Multi-protocol Label Switching (MPLS) – with legacy ATM switching, Lucent has given carriers a vehicle for creating converged multiservice networks that can draw on the revenue of today’s market until IP profits materialize.

The second reason the 880 is important is that it addresses a technical problem with MPLS: MPLS paths look a lot like frame relay or ATM virtual circuits, but they’re not identical. Lucent promises an implementation of MPLS that’s, in Lucent’s words, “connection-oriented,” meaning that it would behave like ATM or frame relay in nearly all respects. Carriers would like to use MPLS for voice and data transport, but they’ve been deterred because MPLS doesn’t assure in-order arrival of packets, provide unambiguous notifications of network faults, support full quality-of-service guarantees or provide virtual circuit levels of security without parallel encryption. Lucent says the 880 can do all this.

But the final reason for the 880’s importance is that it puts MPLS capability into the core and lets MPLS-based services deploy immediately. Yes, that would include IP services, but we know those services aren’t immediately profitable based on our recent experience with the ISPs. What would be really important would be if the marketplace began to deploy services such as frame relay and ATM over MPLS instead of ATM, and MPLS, not ATM, were to form the basis for packet voice networks.

The prospect of MPLS in the network core begs the question of whether MPLS needs to get closer to the user than that. It would be possible to use “connection-oriented” MPLS to create frame relay and ATM services, and support transparent LAN services or wide-area virtual LANs. It also would be possible to continue to use the ATM features of a box such as the 880 to support legacy services and save MPLS as a feature for an IP-dominated world.

Which will it be? It’s an important question because if the next wave of facility-based carrier investment at the edge goes strongly in an ATM/frame relay direction, it will deter MPLS playing a direct revenue-generating role. But keeping the edge ATM-oriented is in the interest of the current ATM incumbents, like Lucent, that stand to gain by reinvestment in legacy edge technology.

Despite this, there’s positive news for MPLS. Start-ups with MPLS edge capabilities, such as WaveSmith Networks Inc. and Equipe Communications Corp., are suddenly in the news. Other MPLS edge players say off the record that they’re attracting carrier interest – and big carriers at that. Will Lucent’s move and the inevitable countermoves of Lucent’s competitors create a real convergence on MPLS?

For users, that might be a good thing. Service features are created at the edge, and healthy competition there could ignite a war among vendors aimed at creating new features that lie in that zone between Level 2 virtual circuits and Level 3 routing. That zone may be the next one carriers plumb for profits and enterprise planners look to for service innovations.

But it’s too soon to tell just what might happen in this area. In the U.S., the regional Bell operating companies are hedging on their edge plans, possibly because the regulatory framework isn’t yet clear. Incumbent equipment vendors are debating whether to emphasize “new MPLS” edge devices or simply modernize frame relay and ATM access technology. With the 880, Lucent has positioned MPLS firmly in the core, but carriers may have to vote with their dollars to resolve the question of MPLS in the edge.

Nolle is president of CIMI Corp., a technology assessment firm in Voorhees, N.J. He can be reached at (856) 753-0004 or